[rocuinneagain]

PyKX converts a q CharVector to Python NumPy array of bytes by default.

https://code.kx.com/pykx/2.5/pykx-under-q/upgrade.html#functional-differences

q).pykx.eval["lambda x: print(type(x))"][test];
<class 'numpy.ndarray'>

q).pykx.eval["lambda x: print(x)"][test];
[b'<' b't' b'a' b'b' b'l' b'e' b'>' b'<' b't' b'r' b'>' b'<' b't' b'h' ...

You can create a helper to convert to strings as you require:

q)b2s:.pykx.eval["lambda x: x.tobytes().decode('UTF-8')"]
q).pykx.print pd[`:read_html]b2s[test]

[  First Name Last Name  Age
0       John       Doe   30
1       Jane     Smith   25
2      Emily     Jones   22]

More info on default conversions:

https://code.kx.com/pykx/2.5/pykx-under-q/intro.html#function-argument-types

And:

https://code.kx.com/pykx/2.5/user-guide/fundamentals/text.html

[rocuinneagain]

@laura – This was a format testing reply – can be deleted. Thanks

[rocuinneagain]

<div>See these sections of docs on:</div>

• List Notation
• Bracket Notation

Order of evaluation also matters: Left-of-Right Evaluation

abs(price)=({abs max x};price)

Will run in order:

1. ({abs max x};price)
2. (price)=({abs max x};price)
   
3. abs(price)=({abs max x};price)
   

whereas:

abs[price]=({abs max x};price)

Will run in order:

1. ({abs max x};price)
2. abs[price]
3. abs[price]=({abs max x};price)

This tries to pass 2 arguments to all which only takes one argument so you get a rank error

https://code.kx.com/q/ref/all-any/#all

all[col=a;col2=b] 

[rocuinneagain]

You could make windows like so:

q)start:0D 
q)duration:1D 
q)length1:0D00:20 
q)length2:0D00:10 
q)windows:{flip (0;y-1)+\:y*til `long$x div y}'[duration;(length1;length2)] 
q)(first;last)@\:windows 0 0D00:00:00.000000000 0D00:19:59.999999999 0D23:40:00.000000000 0D23:59:59.999999999 
q)(first;last)@\:windows 1 0D00:00:00.000000000 0D00:09:59.999999999 0D23:50:00.000000000 0D23:59:59.999999999 
q)t:([] time:100?1D;sym:100?`A`B`C) 
time sym 
------------------------
0D11:37:38.441374450 A 
0D16:23:13.911677598 A 
0D03:40:38.809463381 C 
0D12:50:31.973013281 B 
0D10:58:33.142468035 C 
0D00:11:36.601854264 B

Query for small tables for each sym for each windows:

syms:exec distinct sym from t 
q){[x;y;z] select from x where sym=y,time within z}[t](.)/:syms cross enlist each windows 0 
+`time`sym!(`timespan$();`symbol$()) 
+`time`sym!(,0D03:19:44.361107647;,`A) 
+`time`sym!(0D03:20:29.103351384 0D03:28:31.211699992;`A`A) 
... 
+`time`sym!(,0D20:47:50.484825968;,`B) 
+`time`sym!(,0D21:14:37.927203476;,`B) 
+`time`sym!(0D21:39:42.773356139 0D21:37:00.400066673;`B`B) 
+`time`sym!(`timespan$();`symbol$()) 
+`time`sym!(,0D22:09:45.092769563;,`B)

[rocuinneagain]

The is a kdb+ Solace interface

• https://solace.com/kx/
• https://github.com/KxSystems/solace/tree/master
  
• https://code.kx.com/q/wp/solace/
  

One of the getPayloadAs* functions may be what you need

https://github.com/KxSystems/solace/blob/77dd10edefa014f39b6c5ed9a693bdcad7685509/q/solace.q#L30

q)msg
"\PAR1\025\004\025\010...
q).solace.getPayloadAsBinary msg

[rocuinneagain]

, rather than ; separating the arguments of fbis. Otherwise runs fine:

q)fbis:{[c_;x0;p_]{[xn ;c;p] xn-((xn xexp p)-c)%(p*xn xexp p-1)}[;c_;p_]\[x0]}
q)fbis[2;1;3]
1
1.333333
1.263889
1.259933
1.259921
1.259921

[rocuinneagain]

Linux is best to make use of all features.

If you plan on using Platform/Developer I’d recommend Ubuntu 20.04 for best compatibility.

https://code.kx.com/developer/getting-started/#prerequisites

[rocuinneagain]

Dictionaries require lists to be passed to either side of !. You can use enlist to make lists from atoms:

• https://code.kx.com/q/ref/enlist/
• https://code.kx.com/q/basics/dictsandtables/#construction

q)d:.Q.def[enlist[`num1]!enlist 5].Q.opt .z.x
q)d
num1| 5

If you are using kdb 4.1 you can use the new dictionary syntax which does not need enlist

• https://code.kx.com/q//releases/ChangesIn4.1/#dictionary-literal-syntax

q)d:.Q.def[([num1:5])].Q.opt .z.x
q)d
num1| 5

[rocuinneagain]

You need to install q to use developer. q cannot be installed by pip.

Instructions are here: https://code.kx.com/q/learn/install/

code.kx.com

Installing kdb+ | Learn | kdb+ and q documentation - Kdb+ and q documentation

How to install kdb+ on Linux, macOS, or Windows

[rocuinneagain]

Did you download q from this link?

https://kx.com/kdb-personal-edition-download/

It does not include the cloud libraries.

Instead the insights edition is needed to have those libraries:

https://kx.com/kdb-insights-personal-edition-license-download/

[rocuinneagain]

Node.js is required for the upload server.

Upload – KX Dashboards

If node is not installed, dashboards will work as normal without the upload component.

Documentation will be updated to detail this.

[rocuinneagain]

Going forward PyKX under q is the recommended replacement for embedPy.

• Using PyKX within a q session
• Upgrading from embedPy

*It is not operational yet on Windows

[rocuinneagain]

The first error is saying you do not have node.js installed to be able to run that command

https://nodejs.org/en/download

The second error is saying that a different process is using the port 10001 blocking your new process from starting. You need to kill that other process or change your configuration to use a different port instead.

nodejs.org

Node.js — Download Node.js®

Node.js® is a JavaScript runtime built on Chrome's V8 JavaScript engine.

[rocuinneagain]

You have not exported PATH so q is available on it.

See: https://code.kx.com/q/learn/install/#step-5-edit-your-profile

setx PATH "%PATH%;C:\q\w64"

Or instead to make a permanent change to the existing PATH environment variable on windows you can follow this guide https://chlee.co/how-to-setup-environment-variables-for-windows-mac-and-linux/

code.kx.com

Installing kdb+ | Learn | kdb+ and q documentation - Kdb+ and q documentation

How to install kdb+ on Linux, macOS, or Windows

[rocuinneagain]

Run ‘ls’ on ‘/q’ in the container to see what files are there. The load error output suggests the cloud libraries are not there.